This paper presents the application of a nonlinear controller, using a predictive control strategy to an interconnected tank system. The design procedure of the controller uses the mathematical input-output model of the plant based on the basic physical relations to find a controller output. The applied search strategy minimizes the cost function for a given prediction horizon.

Since the nonlinear model that has been developed is linear in its parameters, it is possible to use the Least Squares method to estimate them. The recursive version is applied in order to model time-varying effects and to track disturbances. Thereby the presented controller forms an adaptive control system. A similar control strategy is possible also for related problems since level control is a frequently occuring problem.